You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, you question is similar to #11. I have given a specific answer to that issue, maybe you can get some help by referring to it.
Besides, for your question, the output box of the model is a normalized coordinate, you just need to know the original image size and multiply back the original length and width correspondence.
哈喽,作者您好,代码中有一些不明白的地方,最后的iou是根据什么计算的,我尝试可视化结果,可是最后模型输出并不能实现定位,定位框所在太小了,或者您可以提供可视化的代码吗
The text was updated successfully, but these errors were encountered: